14 December 2013

–1 0 1 2 3 4 5 6 7 8 9–2 orless

morethan 9

Percent of nonmetropolitan counties or metropolitan areas

U.S. Bureau of Economic Analysis

35

30

25

20

15

10

5

0

Chart 1. Distribution of Personal Income Growth Rates, 2012Chart 1. Distribution of Personal Income Growth Rates, 2012

Metropolitan areasNonmetropolitan counties

Percent change

ERSONAL INCOME grew more slowly (3.7 per-cent) in the nonmetropolitan portion of the

United States in 2012 than in the metropolitan portion(4.2 percent).1 Growth ranged from –33.4 percent inHamilton County, Kansas, to 52.3 percent in WilliamsCounty, North Dakota, but most counties and metro-politan statistical areas (MSAs) grew at rates between2.0 percent and 6.0 percent (chart 1). Inflation, as mea-sured by the national price index for personal con-sumption expenditures, was 1.8 percent in 2012.

The local area personal income estimates presentedin this article continue the successively more detailedseries of data releases from the Bureau of EconomicAnalysis (BEA) depicting the geographic distributionof the nation’s production and income for 2012. Na-tional estimates of personal income and gross domes-tic product (GDP) for 2012 were released in January2013, followed by state personal income estimates inMarch, state GDP estimates in June, and metropolitan

1. Personal income, which is measured in current dollars, is the sum ofnet earnings by place of residence, property income, and personal currenttransfer receipts.

area GDP estimates in September. The local area per-sonal income estimates provide the first glimpse ofpersonal income and compensation by industry innonmetropolitan counties for 2012 and a more de-tailed look at the industrial composition of economicactivity within multicounty MSAs. The geographicpicture of 2012 will be completed with the release ofreal personal income for states and metropolitan areasin April 2014.

The estimates discussed in this article are the resultof the most recent comprehensive revision of the localarea personal income accounts, which was released inNovember 2013. In comprehensive revisions, manifoldimprovements in concepts, definitions, classifications,and statistical methods are introduced into BEA’s eco-nomic accounts to ensure that the accounts continueto accurately describe the evolving American economy.This comprehensive revision incorporated changesthat were adopted as part of the comprehensive revi-sions of the national income and product accounts(NIPAs) and state personal income accounts, whichwere released in July and September 2013, respectively.It also introduced new and updated county-levelsource data as well as certain new data sources thathave never been used before.

This article discusses the patterns and sources of in-come growth for 2012 in nonmetropolitan counties. Itcomplements the discussion of the patterns andsources of production growth for 2012 in metropolitanareas in the October issue of the SURVEY OF CURRENTBUSINESS.2 This article also highlights the fluctuatingboundary between the metropolitan and nonmetro-politan portions of the United States by examining indetail the counties affected by the revised MSA defini-tions released by the Office of Management and Bud-get (OMB) earlier this year. In addition, the articleprovides details about the comprehensive revision oflocal area personal income statistics and summarizesthe major data sources used to prepare the estimates. Abox discusses alternative measures of county wages.

2. See Sharon D. Panek, Jacob R. Hinson, and Frank T. Baumgardner,“Gross Domestic Product by Metropolitan Area,” SURVEY 93 (October2013): 105–141.

Comprehensive Revision of Local Area Personal IncomeNew Statistics for 2012 and Revised Statistics for 2001–2011By David G. Lenze

P

http://www.bea.gov/scb/pdf/2013/10%20October/1013_gpd_by_metropolitan_area.pdf

December 2013 SURVEY OF CURRENT BUSINESS 15

Growth in Nonmetropolitan Counties For statistical purposes, nonmetropolitan counties are those counties that remain after MSAs have been delineated. As defined by OMB, an MSA has at least one urbanized area of 50,000 or more residents plus adjacent territory that has a high degree of social and economic integration with the core as measured by commuting ties. MSAs are defined in terms of whole counties. By these criteria, there are 1,967 nonmetropolitan counties and 1,146 metropolitan counties in the United States.3

Not surprisingly, all nonmetropolitan counties are sparsely populated. They range from Loving County Texas, with a population of 71 and a population density of 0.1 persons per square mile, to Litchfield County Connecticut, with a population of 187,530 and density of 206 persons per square mile.4 The converse is not true; that is, not all metropolitan counties are densely populated. For instance, there are 95 metropolitan counties with a population density below 100 persons per square mile and with less than 30,000 residents, none of whom live in an urbanized area. Evidently, these counties are metropolitan because by the OMB definition, there is a high degree of social and economic integration with the core of an MSA as measured by commuting ties. In other words, these counties are metropolitan because their residents commute to work in another metropolitan county, not because they have an urban character.

The nonmetropolitan portion of the country accounted for slightly less than 10 percent of the nation’s

3. Personal income statistics are available for 3,113 of the 3,143 counties identified by Federal Information Processing Standards (FIPS) codes. BEA combines some small counties (mostly in Virginia but also in Hawaii) with larger nearby counties. For details see the appendix to the Local Area Personal Income Methodology available on the BEA Web site.

earnings in 2012, but reflecting the rural affinity of much mining and farming, the nonmetropolitan portion of the United States accounted for almost 37 percent of national earnings in natural resource industries (table A). The nonmetropolitan area also accounted for 14.7 percent of manufacturing earnings, 12.6 percent of transportation earnings, and 12.1 percent of government earnings. In contrast, relatively little—less than 4.0 percent—of earnings in the information, finance, and business services industries was generated in nonmetropolitan counties.

Not only did the nonmetropolitan portion grow more slowly than the metropolitan portion of the United States in 2012, its 3.7 percent personal income growth rate was a substantial slowdown from its 6.4 percent growth in 2011 (table B). Most of the slowdown was attributable to net earnings, which grew

Table A. Industrial Structure of Metropolitan and Nonmetropolitan Portions of the United States for 2012

Earnings by place of work

(billions of dollars)

Industry’s share of area’s total earnings

(percent)

Nonmetropolitan

share of national earnings (percent)

Metropolitan

Nonmetropolitan

Metropolitan

Nonmetropolitan

Natural resources 1................................... 187.0 108.9 2.1 11.3 36.8 Construction ............................................ 460.9 56.5 5.2 5.9 10.9 Manufacturing .......................................... 829.2 142.9 9.4 14.9 14.7 Wholesale and retail trade ....................... 986.6 102.3 11.1 10.6 9.4 Transpor tation, warehousing, and utilities 360.0 52.0 4.1 5.4 12.6 Information ............................................... 303.5 10.2 3.4 1.1 3.3 Finance and insurance ............................ 664.5 26.3 7.5 2.7 3.8 Real estate and rental and leasing .......... 171.1 10.2 1.9 1.1 5.6 Business services 2 ................................. 1,562.4 61.6 17.6 6.4 3.8 Education, health care, and social

assistance ............................................ 1,133.8 105.9 12.8 11.0 8.5 Leisure, hospitality, and other 3................. 690.8 76.8 7.8 8.0 10.0 Government and government enterprises 1,510.9 207.1 17.1 21.6 12.1

Local ................................................. 791.2 124.2 8.9 12.9 13.6

Total......................................................... 8,860.7 960.7 100.0 100.0 9.8

4. Population densities are from the Census Bureau’s American Factfinder 1. Consists of farm; forestry, fishing, and related activities; and mining. and refer to 2010; population refers to 2012. The Census Bureau uses two 2. Consists of professional, scientific, and technical services; management of companies and enterprises; and population density thresholds in the delineation of urban areas: 1,000 per- administrative and waste management services.

3. Consists of arts, entertainment and recreation; accommodation and food services; and other services, sons per square mile and 500 persons per square mile. except public administration.

Table B. Personal Income Change by Component for U.S. Metropolitan and Nonmetropolitan Portions

Percent change Contribution to percent change

in personal income (percentage points)

Dollar change (millions of dollars)

Personal income

Net earnings 1

Dividends, interest, and rent

Transfer receipts

Net earnings 1

Dividends, interest, and rent

Transfer receipts

Personal income

Net earnings 1

Dividends, interest, and rent

Transfer receipts

2010–2011 United States .......................................................... 6.1 6.2 10.6 1.3 4.0 1.8 0.2 756,229 500,134 226,078 30,017

Metropolitan por tion .............................................. 6.0 6.1 10.6 1.3 4.0 1.8 0.2 662,027 437,056 200,091 24,880 Nonmetropolitan por tion ....................................... 6.4 7.5 10.3 1.3 4.3 1.8 0.3 94,202 63,078 25,987 5,137

2011–2012 United States .......................................................... 4.2 4.3 5.5 2.2 2.8 1.0 0.4 549,502 367,526 130,630 51,346

Metropolitan por tion .............................................. 4.2 4.4 5.5 2.3 2.9 1.0 0.4 490,611 332,244 114,712 43,655 Nonmetropolitan por tion ....................................... 3.7 3.9 5.7 2.0 2.2 1.0 0.5 58,891 35,282 15,918 7,691

1. Earnings by place of work net of contributions for government social insurance and net of the residence adjustment.

http://www.bea.gov/regional/pdf/lapi2011.pdfhttp://www.bea.gov/regional/pdf/lapi2011.pdf

16

Local Area Personal Income December 2013

only 3.9 percent in 2012, down from 7.5 percent in 2011. Property income (dividends, interest, and rent), which accounted for 18 percent of personal income, also grew substantially slower (5.7 percent) in 2012 than in 2011 (10.3 percent) in nonmetropolitan counties. Transfer receipts, which accounted for 24.5 percent of personal income, accelerated slightly to 2.0 percent growth in 2012 from 1.3 percent in 2011.

Earnings growth by industry provides additional insight into the reasons for these disparities in personal income growth rates (chart 2 and table C). Farm earnings for the United States fell 1.2 percent in 2012 after growing 38.9 percent in 2011. Like farming, government earnings growth was weak in 2012, growing 0.6 percent, but unlike farming, government earnings growth was also weak in 2011, growing 0.4 percent. In contrast to farming and the public sector, manufacturing earnings grew at a much higher, 4.3 percent rate in 2012, after growing 5.6 percent in 2011.

Per capita personal income (personal income divided by population) in nonmetropolitan counties in 2012 ranged from $116,978 in Williams County, North Dakota, to $17,922 in Telfair County, Georgia. Net earnings was the source of most of the income in Williams County and amounted to $100,138 per person. Mining (which includes oil and gas extraction) accounted for 43 percent of earnings in Williams County.

In contrast, per capita personal income in Telfair County has stagnated and in 2012 was 2.4 percent below the level for 2001. Net earnings per person in Tel-fair County was only $8,324. Transfer receipts were

CharChartt 3.3. PPeer Capita Pr Capita Peerrssonal Incomeonal Income

2001 02 03 04 05 06 07 08 09 10 11 2012

Dollars per person

U.S. Bureau of Economic Analysis

120,000

100,000

80,000

60,000

40,000

20,000

0

Williams County, North Dakota Telfair County, Georgia U.S. average

Table C. Growth of U.S. Earnings by Industry

The high level of per capita personal income in Wil- Dollar changePercent change (millions of dollars)liams County is a recent phenomenon. As recently as 2007, per capita personal income in the county was 2011 2012 2011 2012 slightly below the national average (chart 3). By 2012, Private ................................................................ 5.9 5.1 426,595 389,652 it was more than twice as large. Farm ............................................................... 38.9 –1.2 28,278 –1,250

Nonfarm .......................................................... 5.5 5.1 398,317 390,902 Forestry, fishing, and related activities ........ –2.9 8.2 –756 2,112 Mining.......................................................... 38.0 11.7 41,494 17,623CharChartt 2.2. UU.S..S. Earnings bEarnings byy SectorSector Utilities......................................................... 5.8 0.0 4,335 35 Construction ................................................ 3.0 6.4 14,381 31,124 Manufacturing.............................................. 5.6 4.3 49,374 40,056

Durable goods manufacturing.................. 6.2 4.7 34,555 27,616 Nondurable goods manufacturing............ 4.5 3.6 14,819 12,440

Wholesale trade .......................................... 6.2 5.5 27,863 26,151 Retail trade.................................................. 3.9 3.6 21,309 20,230 Transportation and warehousing ................. 8.4 5.4 24,436 16,919 Information .................................................. 4.2 4.3 12,225 12,822 Finance and insurance................................ 1.0 3.0 6,591 19,894 Real estate and rental and leasing.............. 18.6 6.4 26,789 10,837 Professional, scientific, and technical

services ................................................... 7.2 6.4 61,568 58,747 Management of companies and enterprises 7.0 9.0 15,344 21,253 Administrative and waste management

services ................................................... 7.2 6.1 24,877 22,472 Educational services ................................... 3.8 4.9 5,744 7,736 Health care and social assistance............... 3.0 4.1 29,773 42,133 Arts, entertainment, and recreation............. 3.7 4.9 3,558 4,952 Accommodation and food services ............. 6.8 7.2 18,297 20,533 Other services, except public administration 3.4 4.5 11,115 15,273

Government and government enterprises.......... 0.4 0.6 7,588 10,243 Federal, civilian ............................................... 2.4 –0.3 6,963 –815 Military ............................................................ –0.6 –0.7 –902 –1,001 State government ........................................... 1.2 1.6 4,186 5,616 Local government ........................................... –0.3 0.7 –2,659 6,443

Total....................................................................... 4.8 4.2 434,183 399,895

2002 03 04 05 06 07 08 09 10 11 2012

Percent change

U.S. Bureau of Economic Analysis

50

40

30

20

10

0

–10

–20

–30

Farm Government and government enterprises Private nonfarm

17 December 2013 SURVEY OF CURRENT BUSINESS

also an important source of personal income in 2012, amounting to $6,862 per person. The low level of per capita personal income in Telfair County reflects a relatively large proportion of the population (20.6 percent) living in group quarters with little income, including the inmates of a state prison.

Revised Metropolitan

Statistical Area (MSA) Definitions

The OMB revised its definitions of metropolitan statistical areas in February 2013. In doing so, it designated 23 new MSAs, merged 5 separate MSAs with adjacent MSAs, and changed 3 MSAs to micropolitan status (the new MSAs are highlighted in chart 4).5 The net result was to raise the number of MSAs to 381. In the process of redefining the MSAs, 101 counties converted from nonmetropolitan status to metropolitan, and 36 metropolitan counties reverted from metropol

5. The three MSAs that changed to micropolitan statistical areas are (1) Danville, VA; (2) Holland-Grand Haven, MI; and (3) Sandusky, OH. The five MSAs that merged with other MSAs are (1) Palm Coast, FL, which is now part of the Deltona-Daytona Beach-Ormond Beach, FL MSA; , (2)Pascagoula, MS, which is now part of the Gulfport-Biloxi-Pascagoula, MS MSA; (3) Poughkeepsie-Newburgh-Middletown, NY, which is now part of the New York-Newark-Jersey City, NY-NJ-PA MSA; (4) Anderson, IN, which is now part of the Indianapolis-Carmel-Anderson, IN MSA; and (5) Anderson, SC, which is now part of the Greenville-Anderson-Mauldin, SC MSA.

Chart 4. Metropolitan and Nonmetropolitan Areas

itan to nonmetropolitan status. This raised the metropolitan population of the United States 1.1 percent and reduced the nonmetropolitan population 5.7 percent.

When OMB redefines its MSAs, BEA rebuilds the entire time series for the MSAs so that the local area data in the interactive tables on the BEA Web site use the same definition for every year in the time series. This is easily done because MSAs are defined in terms of counties. For example, when OMB first defined the Gainesville, FL MSA it consisted of the single county of Alachua. The current definition of the Gainesville, FL MSA consists of Alachua and Gilchrist counties. BEA’s estimates of personal income for the Gainesville, FL MSA also consist of the same two counties for every year that data are available.

The populations of the new MSAs range from 63,399 residents to 193,882 residents. Evidently, they have crossed the demographic threshold required for metropolitan status, but in some respects their average incomes and industrial composition continue to resemble nonmetropolitan areas.

On average, per capita personal income in the new MSAs was $37,165 in 2012, only 5.2 percent above the $35,324 nonmetropolitan average and 17.8 percent below the $45,188 metropolitan average. Seven of the new MSAs have per capita personal incomes below the

New metropolitan areas

Other metropolitan areas

Nonmetropolitan area

U.S. Bureau of Economic Analysis

18 Local Area Personal Income December 2013

nonmetropolitan average, and two have per capita incomes above the metropolitan average.

As noted above, metropolitan areas tend to have relatively large professional services, finance, and information industries, compared with nonmetropolitan areas. However, among the new MSAs, only California, MD, had a professional services industry as large as the metropolitan average in 2012.6 Professional services accounted for 26.4 percent of earnings in California, MD, compared with 10.6 percent for the metropolitan average. Only Hammond, LA, had a finance sector as

6. The magnitudes of the professional services industry in Hilton Head Island, SC, and in The Villages, FL, are unknown because of nondisclosure rules.

Data Availability All of the local area personal income data presented in this article, along with much additional detail, are available in interactive data tables on the BEA Web site. Data are available for counties, metropolitan statistical areas (MSAs) and other combinations of counties at www.bea.gov.

The data for 2001–2012, the years covered by the North American Industrial Classification System (NAICS), have been revised to be consistent with the comprehensive revisions of the national income and product accounts and the state personal income accounts. Data for 1969–2000, the years covered by the Standard Industrial Classification (SIC), are scheduled to be revised in the spring of 2014. Unrevised data for 1969–2000 remain in the interactive tables as a convenience, but users are advised that these data are not comparable with the more recent estimates.

The impact of sequestration and reduced fiscal year 2013 funding levels for the Bureau of Economic Analysis (BEA) have required reductions in the Bureau's local area personal income (LAPI) program. Effective with this release, the following statistical detail will not be updated or made available: (1) local area employment by industry; (2) detailed statistics on personal current transfer receipts; (3) detailed statistics on farm income and expenses; and (4) statistics for BEA Economic Areas. In addition, industry detail on compensation and earnings has been reduced from 108 industries to 25 industries. The loss of statistical detail has a significant effect on the interactive data tables available to the public. For an explanation of the specific LAPI tables eliminated or modified by sequestration and reduced fiscal year 2013 funding levels, please see: www.bea.gov/_pdf/ sequestration_fact_sheet_with_appendix.pdf.

For further information about the statistics, contact the Regional Income Division at 202–606–5360, or e-mail reis@bea.gov.

large as the metropolitan average. Finance accounted for 8.3 percent of earnings in Hammond, LA compared with 7.5 percent for the metropolitan average. And only Staunton, VA, had an information sector as large as the metropolitan average.7 Information accounted for 4.0 percent of earnings in Staunton, VA, compared with 3.4 percent for the metropolitan average.

On the other hand, nonmetropolitan areas tend to have relatively large farming, manufacturing, and government sectors. Farming in Grand Island, NE (13.2 percent of earnings) and Walla Walla, WA (11.1 percent) exceeded the 6.0 percent average for nonmetropolitan counties in 2012. Manufacturing in Midland, MI (26.6 percent), Albany, OR (23.9 percent), Gettysburg, PA (19.4 percent), Chambersburg, PA (18.5 percent), Staunton, VA (18.0 percent), Grand Island, NE (17.0 percent), and East Stroudsburg, PA (15.0 percent) exceeded the 14.9 percent average for nonmetropolitan counties. Government earnings exceeded the nonmetropolitan average of 21.6 percent in 10 of the new MSAs.

Comprehensive Revision On November 21, 2013, BEA released the initial results of its latest comprehensive, or benchmark, revision of the local area personal income statistics; the results of the previous comprehensive revision were released in April 2010.8

The first installment of the 2013 revision, consists of new and revised statistics for the years covered by the North America Industry Classification System (NAICS); that is, from 2001 through 2012. Additional revisions, covering 1969–2000 for the years covered by the Standard Industrial Classification (SIC) are scheduled to be released in the spring of 2014.

Especially noteworthy in the 2013 comprehensive revision was the introduction of county-level data to improve the estimates of the Medicare benefits and Supplemental Nutritional Assistance Program benefits, two components of personal current transfer receipts.

The 2013 local area personal income comprehensive revision incorporated the changes that were adopted as part of the comprehensive revisions of the national income and product accounts (NIPAs), which was released in July 2013 and of the state personal income accounts which were released in September.9

7. The magnitude of the information industry in New Bern, NC, is unknown because of nondisclosure rules.

8. See David G. Lenze, “Comprehensive Revision of Local Area Personal Income,” SURVEY 90 (May 2010): 22–30.

9. See Robert Kornfeld, “Initial Results of the 2013 Comprehensive Revision of the National Income and Product Accounts,” SURVEY 93 (August 2013): 6–17 and David G. Lenze, “Regional Quarterly Report: Comprehensive Revision,” SURVEY 93 (November 2013): 48–58.

mailto:reis@bea.govwww.bea.gov/_pdfhttp://www.bea.govhttp://www.bea.gov/_pdf/sequestration_fact_sheet_with_appendix.pdfhttp://www.bea.gov/_pdf/sequestration_fact_sheet_with_appendix.pdfhttp://www.bea.gov/scb/pdf/2010/05%20May/0510_lapi-text.pdfhttp://www.bea.gov/scb/pdf/2010/05%20May/0510_lapi-text.pdfhttp://www.bea.gov/scb/pdf/2013/08%20August/0813_nipa-revision%20text.pdfhttp://www.bea.gov/scb/pdf/2013/08%20August/0813_nipa-revision%20text.pdfhttp://www.bea.gov/scb/pdf/2013/11%20November/1113_regional_report.pdfhttp://www.bea.gov/scb/pdf/2013/11%20November/1113_regional_report.pdf

19 December 2013 SURVEY OF CURRENT BUSINESS

The rest of this section will describe briefly the magnitude of the revisions and then describe the improvements in source data and statistical methods.

Magnitude of revisions For many counties, the picture of personal income shown by the revised estimates is similar to the picture shown by the previous estimates (table D). More than 80 percent of the revisions in every year were less than 5 percent in absolute value. For example, in 2001, 41 percent of the revisions to county personal income were less than 1 percent and 57 percent were between 1 percent and 5 percent. Only 56 of the 3,110 counties were revised 5 percent or more. For the most recent years, there were more large revisions—143 counties were revised 10 percent or more in 2011—but this of-

Table D. Revisions to County Personal Income, 2001–2011

Revision (absolute value)

Number of counties

2001 2002 1 2003 2004 2005 2006 2007 2008 2 2009 3 2010 2011

0.0–0.9 percent ........ 1.0–4.9 percent ........ 5.0–9.9 percent ........ 10.0 percent or more

Total .........................

1,289 1,765

50 6

3,110

1,361 1,686

55 9

3,111

1,158 1,823

121 9

3,111

894 2,082

127 8

3,111

794 2,156

147 14

3,111

741 2,158

191 21

3,111

937 2,015

140 19

3,111

1,346 1,611

126 29

3,112

635 2,195

245 38

3,113

883 1,976

218 36

3,113

662 1,890

418 143

3,113

1. For 2002 forward, the number of counties includes Broomfield County, CO. 2. For 2008 forward, the number of counties reflects the division of Skagway-Hoonah-Angoon Census Area

into the Skagway Borough and the Hoonah-Angoon Census Area. 3. For 2009 forward, the number of counties reflects the division of the Wrangell-Petersburg Census Area into

the Petersburg Census Area and the Wrangell City and Borough.

ten reflects the replacement of preliminary estimates of certain components of personal income based on simple extrapolations with estimates based on recently released source data.

Medicare benefits Previously, county estimates of Medicare benefits were extrapolated from a benchmark set of estimates based on reimbursements for hospital and medical expenses by county for 1995 from the Centers for Medicare and Medicaid Services (CMS). The 1995 estimates were extrapolated to the present by the change in Medicare enrollment (as of July of each year), also from CMS. As part of the comprehensive revision, the estimates of Medicare benefits are now based on fee-for-service per capita expenditure data by county from CMS. These data, which are available annually, are combined with annual Medicare enrollment by county to obtain an estimate of total Medicare benefits.

Supplemental Nutritional Assistance Program (SNAP) benefits The basic local area data source for SNAP benefits are payments data from various state departments of social services. Formerly, when payments data were not available, the county distribution of benefits was held constant. As part of the comprehensive revision, the number of benefit recipients by county from the U.S. Census Bureau were used in the allocation of the state total when county payments data were unavailable.

Acknowledgments The Regional Income Division of the Bureau of Eco- social insurance, and the adjustment for residence. Major nomic Analysis (BEA), under the direction of Mauricio responsibilities were assigned to Brian J. Maisano, Lisa C. Ortiz, Chief, prepared the annual estimates of local area Ninomiya, James P. Stehle, and Matthew A. von Kerczek. personal income. Joel D. Platt, Associate Director for Contributing staff members were Suet M. Boudhraa, Regional Economics, provided general guidance. The Andy K. Kim, Toan A. Ly, W. Timothy McKeel, Linda M. preparation of the revised estimates was a division-wide Morey, Anand N. Seeram, and Troy P. Watson. effort. The Farm Income and Employment Section, under the

The Compensation Branch, under the supervision of supervision of James M. Zavrel, Assistant to the Division John A. Rusinko, Chief, prepared the estimates of non- Chief, prepared the estimates of farm wages and salaries, farm wages and salaries, supplements to wages and sala- farm supplements to wages and salaries, and farm prories, and personal current tax receipts. Major prietors’ income. Major responsibilities were assigned to responsibilities were assigned to Peter Battikha, Michael Carrie L. Litkowski. Contributing staff members were L. Berry, Elizabeth P. Cologer, John D. Laffman, David G. Daniel R. Corrin and Michelle A. Harder. Lenze, Paul K. Medzerian, and Joseph L. Stauffer. Con- The Data and Administrative Systems Group assemtributing staff members were Susan P. Den Herder, Ter- bled the public use tabulations and data files and preence J. Fallon, Michael W. Jadoo, Russell C. Lusher, pared the tables. Major responsibilities were assigned to Nathaniel R. Milhous, Michael A. Reid, and Ross A. Jeffrey L. Newman, Michael J. Paris, and Callan S. Swen-Stepp. son. Contributing staff members were Brooke N. Huo-

The Regional Income Branch prepared the estimates of tari, Monique B. Tyes, Melanie N. Vejdani, and Jonas D. nonfarm proprietors’ income, property income, personal Wilson. current transfer receipts, contributions for government

20 Local Area Personal Income December 2013

Personal contributions for veterans life insurance Formerly, state estimates of personal contributions for veterans life insurance (a component of contributions for government social insurance) were allocated to counties in proportion to the veteran population from the Census of Population. The veteran population for 2000 was held constant for all subsequent years. As part of the comprehensive revision, state estimates of contributions for veterans’ life insurance were allocated to counties using the 2006–2010 American Community Survey “5‐year” estimates of the veteran population, centered on 2008, the midpoint of the estimation interval. This allocator will be held constant for subsequent years until new, nonoverlapping 5-year estimates are available.

Home Affordable Mortgage Program principal reduction This recently enacted federal program, a response to the subprime mortgage crisis, helps eligible home owners with loan modifications on their home mortgage debt. In lieu of direct data on benefits, the state estimates are allocated to counties on the basis of Federal Reserve Bank of New York data on the number of mortgage debtors, per debtor mortgage debt balance and percent of mortgage debt in delinquency.

Miscellaneous components of personal income Formerly, the state estimates of a number of small components of personal income for which no county-level source data are available (for example, temporary disability benefits, a component of personal current transfer receipts) were allocated to counties on the basis of civilian population. As part of the comprehensive revision, these components are now allocated to counties using household population, that is, total population excluding persons living in group quarters such as prisons. State estimates of the recently enacted Temporary High Risk Health Insurance premium reduction are also allocated to counties on the basis of household population.

Residence adjustment Estimates of wage and salary flows across the borders of the United States were substantially revised in 2011 as part of the annual revision of the international transactions accounts.10 Because of the magnitude of the revisions and the number of years affected, the introduction of these revised national estimates into the regional personal income accounts was delayed until

10. See Mai-Chi Hoang and Erin M. Whitaker “Annual Revision of the U.S. International Transactions Accounts,” SURVEY 91 (July 2011): 58.

the comprehensive revision. These wage flows are part of the residence adjustment in the local area personal income accounts. They account for wage and salary flows between Canada, Mexico, and the United States. In addition, they account for the inflows of wages and salaries earned by U.S. residents employed by certain international organizations (such as the United Nations, the International Monetary Fund, and the World Bank) and by foreign embassies and consulates located within the geographic borders of the United States.

Changes in statistical methods There were also several statistical improvements to the local area personal income accounts. Some of these improvements (such as for employer contributions for pensions and health insurance) involve state and national source data that are not available for individual counties. However, these improvements are implicitly incorporated into the county estimates, which must sum to the state and national estimates.

Source Data The primary 2012 county-level data used by BEA to prepare the estimates of local area personal income presented in this article were wage and salary data from the Bureau of Labor Statistics, benefits paid by the Social Security Administration, Medicare enrollment and fee-for-service expenditure data from the Centers for Medicare and Medicaid Services, and Medicaid payments from state departments of social services. In addition, tabulations of 2011 federal income tax returns from the Internal Revenue Service were used, primarily for dividends, interest, nonfarm proprietors’ income, and the residence adjustment.11

Other 2012 county-level data used by BEA to prepare estimates of various components of local area personal income include the following:

● Farm cash receipts, government payments, crop production, and livestock inventories by county for 2012 from the U.S. Department of Agriculture were used in the estimation of local area farm income.

● The number of full-time military and coast guard personnel by county for 2012 from the Departments of Defense and Homeland Security was used in the estimation of military earnings.

● County-level data for 2012 from the Federal Assistance Award Data System were used to prepare estimates of some components of personal current transfer receipts.

● Household population by county for 2012 from the Census Bureau was used to allocate state estimates of a few small components of personal income.

11. For complete details about the estimation methodology and data sources, see Local Area Personal Income Methodology on BEA’s Web site.

http:adjustment.11http:accounts.10http://www.bea.gov/scb/pdf/2011/07%20July/0711_ita-annual.pdfhttp://www.bea.gov/scb/pdf/2011/07%20July/0711_ita-annual.pdfhttp://www.bea.gov/regional/pdf/lapi2011.pdf

21 December 2013 SURVEY OF CURRENT BUSINESS

Alternative Measures of County Employment and Wages Three widely used measures of county employment and wages by place of work are (1) employment and payroll in the County Business Patterns (CBP) series from the Census Bureau, (2) employment and wages from the Quarterly Census of Employment and Wages (QCEW) program from the Bureau of Labor Statistics (BLS), and (3) wage and salary disbursements and employment from the Bureau of Economic Analysis (BEA). These measures differ in source data and coverage.

The CBP data are derived from Census Bureau business establishment surveys and federal administrative records. The QCEW data are tabulations of monthly employment and quarterly wages of workers who are covered by state unemployment insurance programs or by the unemployment insurance program for federal employees.1 The BEA estimates of employment and wages are primarily derived from the BLS data; the estimates for industries that are either not covered or not fully covered in the QCEW are also based on supplemental data from other agencies, such as the Department of Defense, the U.S. Department of Agriculture, and the Railroad Retirement Board.

The coverage of the Census Bureau data differs from that of the BLS data primarily because the Census Bureau data exclude most government employees and because the BLS data cover civilian government employees.2 The CBP data also exclude several private industries that are partly covered by the QCEW: crop and animal production; rail transportation; insurance and employee benefit funds; trusts, estates, and agency accounts; and private households. However, the CBP data cover the employees of educational institutions, membership organizations, and small nonprofit organizations in other industries more completely than the BLS data.3 In addition, the Census Bureau reports employment only for the month of March; the BLS employment data are quarterly and annual averages of monthly data.

In 2001, both BLS and BEA began to include employees of Indian tribal councils in local government. These employees were previously included in the relevant private industries.4 In the Census Bureau data, these employees are still classified in private industries.

BEA estimates of employment and wages differ from the BLS

1. The QCEW data account for 93 percent of BEA’s wages and salaries. 2. The Census Bureau data cover only those government employees who

work in government hospitals, federally chartered savings institutions and credit unions, liquor stores, and wholesale liquor establishments, and university publishers. The BLS data in most states exclude state and local elected officials, members of the judiciary, state national and air national guardsmen, temporary emergency employees, and employees in policy and advisory positions.

3. The BLS data do not cover certain religious elementary and secondary schools because a Supreme Court decision exempts some of these schools from unemployment compensation taxes. The BLS data also exclude college students (and their spouses) who are employed by the school in which they are enrolled and student nurses and interns who are employed by hospitals as part of their training. In half of the states, the BLS data only include nonprofit organizations with four or more employees during 20 weeks in a calendar year.

4. For example, employees of casinos owned by tribal councils were included in “Amusement, Gambling, and Recreation Industries.”

data because BEA adjusts the estimates to account for employment and wages that are not covered or not fully covered by the unemployment insurance programs. BEA adds estimates of employment and wages to the BLS data to bridge small gaps in coverage for nonprofit organizations that do not participate in the unemployment insurance program (in several industries), for students and their spouses employed by colleges or universities, for elected officials and members of the judiciary, for interns employed by hospitals and by social service agencies, and for insurance agents classified as statutory employees. In addition, BEA uses supplemental source data to estimate most, or all, of the employment and wages for the following: farms, farm labor contractors and crew leaders, private households, private elementary and secondary schools, religious membership organizations, rail transportation, and military. BEA also adjusts for employment and wages subject to unemployment insurance but not reported by employers. Other adjustments to wages include estimates for unreported tips, judicial fees paid to jurors and witnesses, compensation of prison inmates, and marriage and license fees paid to justices of the peace.5

The Census Bureau released 2011 data for total employment and payrolls for counties on its Web site on April 2013. BLS released county data on total employment and average weekly pay for 2012 on its Web site on September 9, 2013. BEA released preliminary estimates for 2012 and revised estimates for 2010–2011 of total wage employment and total wage and salary disbursements for counties on its Web site on November 21, 2013.

5. For a detailed description of the sources and methods used to prepare the estimates, visit www.bea.gov/regional/methods.cfm.

National Totals of BEA County Estimates of Wages and

Salaries and CBP Payrolls and QCEW Wages

[Billions of dollars]

2010 2011 2012

Total CBP payrolls..................................................... Plus: Differences in coverage:

4,941.0 5,164.9 n.a.

QCEW civilian government wages 1................... 1,031.6 1,033.7 n.a. Other differences, net 2...................................... 3.1 18.7 n.a.

Equals: Total QCEW wages ...................................... Plus: BEA adjustments:

For unrepor ted wages and unrepor ted tips on

5,975.7 6,217.3 6,490.6

employment tax returns.................................. For wages and salaries not covered or not fully

covered by unemployment insurance:

69.7 78.1 80.8

Private......................................................... 194.6 205.4 217.5 Government ................................................ 131.6 130.7 131.0

Other BEA adjustments 3................................... –2.9 –3.2 –2.8 Equals: BEA estimates of wages and salaries 4 ....... 6,368.6 6,628.3 6,917.2

n.a. Not available 1. Adjusted to remove the wages of Indian tribal councils that are included in the Census Bureau’s total

payroll data. 2. Includes differences of coverage in private education, membership organizations, and government. 3. Adjusted to remove wages and salaries of employees of U.S. companies stationed overseas and to

reflect updates to QCEW data. 4. Consists of the earnings of persons who live in the United States and of foreign residents working in

the United States. The regional total differs from the national estimate; see “Personal income in the NIPAs and State Personal Income,” SUR VEY OF CURRENT BUSINESS 93 (November 2013): 57.

NOTE. Details may not equal totals due to rounding.

Michael Jadoo

Tables 1 and 2 follow.

www.bea.gov/regional/methods.cfm

Comprehensive Revision of Local Area Personal Income: New Statistics for 2012 and Revised Statistics for 2001–2011Growth in Nonmetropolitan CountiesRevised MetropolitanStatistical Area (MSA) DefinitionsComprehensive RevisionMagnitude of revisionsMedicare benefitsSupplemental Nutritional Assistance Program (SNAP) benefitsPersonal contributions for veterans life insuranceHome Affordable Mortgage Program principal reductionMiscellaneous components of personal incomeResidence adjustmentChanges in statistical methods

Source DataBoxesBox: Data AvailabilityBox: AcknowledgmentsBox: Alternative Measures of County Employment and Wages

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice